Time: started 8:34 AM Saturday, completed 10:59 AM

Weather: low 14.9c (58.8F) high of 29.5c (85.5F), hot and dry (7% humidity) with cooling mid-afternoon breezes

Crew Physical Status: Pushing to the limit.

Narrative of Field Mission Results: Our final test day had arrived, and we were ready. The crew from JSC arrived at Lith Canyon early to move the robot to the canyon the floor. Maarten, Abby, and Brent labored to produce a formal plan for the Segment 3 EVA, the walk down the canyon to the east and north. I wrote a long report about Thursday, but never caught up with photos.

The balance of the crew left for Lith about 11:15 AM, bringing computers, cameras, coolers, water, and lunch. The weather was pleasant: 80F and 7% humidity. Regular breezes 5 mph from the northwest quickly dropped the temperature to 77F, followed moments later by a still warmth that popped it back to the 80s.

On arrival we realized that the system configuration was again unique and not all the details had been thought out. Now the ERA was in the canyon about .3 km north of us, far out of sight. The Rocky-C ATV with the network relay from the far point repeater had been moved to provide coverage in the canyon to the ERA. The laptop (M-PAK C) running the Mobile Agents system for the ERA was on Rocky-C (added the previous day to decrease the load on the ERA's computers). The astronaut computer backpacks were on the table under the canopy, and now beyond reach of the network. So Ron could ping the ERA from MDRS and M-PAK-C, but couldn't see the backpacks. To start Mobile Agents, we needed all four platforms to be on the network about the same time. Previously, we did this by directly connecting the backpacks to the hub in the ERA's truck next to the canopy (and running M-PAK-C in the truck). Now the truck's hub was offline.

After some consideration, we realized we could set up a battery-powered wireless access point on a rise between the canopy and the Rocky-C ATV. Alternatively, we could carry the backpacks to another canopy we had set up in the canyon near the ERA. But then we'd have to carry the generator, too, because we wanted the start-up process to be on external power. The second option was less risky, but required a lot more physical work. The group voted for setting up the extra access point, and by 12:30 PM the network had been extended to include the canopy area.

We proceeded to set up the machines and load the software. Here is a rough log of events:

• 1 PM Ron reports from MDRS that he's unable to ping M-PAK-A (Brent's backpack, the InfoPack).
• 1:02 PM A stray dust devil comes out of nowhere and whirls through the canopy--we're eating dust.
• 1:20 PM We have reset the access point and all is well again. John checks the microphones.
• 1:22 PM Ron starts HabCom (at MDRS upper deck) and tells Maarten to start the Astronaut backpack computers (M-PAK-A and B).
• 1:28 PM Ron reports HabCom froze; we need to restart all the models on the five computers (ERA, M-PAK-A, B, C, HabCom)
• 1:31 PM Ron's lost communications to Rocky-C ATV (location of M-PAK-C), but it returns.
• 1:42 PM Ron is loading the EVA plan on HabCom; an agent will push it to the other computers.
• 1:44 PM The plan is loaded; Ron reports we should start suiting up the astronauts.
• 1:45 PM We take the backpacks off external power; we now have about 1 hr 45 min before needing to swap batteries in the canyon.
• 1:55 PM Abby is set to go; Brent is putting on his helmet.
• 2:02 PM GPS and Nonin (Bionsensors) started for Abby; Ron does the same for Brent (again operating from MDRS).
• 2:07 PM Abby connects her camera and Ron verifies that the COM port is correct; Brent does the same.

Brent and Abby stepped out from the canopy and began the initialization sequence, commanding their personal agent to track their GPS location and bionsensors. One of them announces the activity of walking to the ERA (using the waypoint number)-the EVA has begun.

About 2:22 PM we arrive down the cow path, on the canyon floor, and Ron reports immediately that he has no communications with the astronauts at the hab. We determine that the computers have not properly reassociated (network roaming) with the ERA access point. Because Ron has no access, we need to handle this locally, removing the backpack covers, flipping open the laptops, and reconfiguring the selected network. We buzz around each other: "Didn't this work before?" and "What's different here?" But the problem was solved and we moved on.

The rest of the EVA was the usual mixture of delight, amazement, and "Hey, what's going on?" The ERA was successfully commanded to watch Brent, and we could easily see the front camera spinning around as he moved. Abby and Brent named locations, took samples and photographs, made voice notes and associations. The data was stored in ScienceOrganizer in the hab and mirrored to Ames (providing access to the RST). We all received 130 emails, sent from 1:42 - 5:05 PM, indicating each record and alerts.

The astronauts successfully commanded the ERA to follow them, as they proceeded to two more waypoints. We had some close calls with overhanging walls and sharp rocks, but generally the robot found its way around and over obstacles without needing to be closely led. The robot took photographs and panoramas as requested, all while Kim and Jeff watched the EVA back in MDRS as the robot visually tracked Brent.

We had the usual hiccups along the way: Brent's backpack overheated at 3:30 PM; the internal temperature was 131F. A second shut down occurred at 3:49 PM with an internal temperature of 116F, so we simply swapped his batteries and removed the cover. Abby's unit had no problems. The obvious difference is that Brent was carrying a second computer inside for the differential GPS, plus perhaps the fans and openings were not as well-placed as on Abby's backpack. (If we can port the dGPS server program from Linux, we could eliminate that extra machine.)

The second trick was maintaining the network, a key aspect of the day's test. Rick had to move the ATV along the ledge (simulating a robot or perhaps a different kind of network). When the astronauts & ERA reached the point around the bend, where our tests indicated that the robot had to be carefully placed in line of sight of the ATV and not moved any further (due to a hill that would obstruct all visibility north), the network became "ratty" and communications to the hab were intermittent. Meanwhile, the astronauts went about their business, and data were transmitted when the network allowed.

Eventually, we instructed the astronauts to guide the ERA further down the canyon, but to a point where relay was improved. Here I noticed that Abby was out of sight of everyone, working once up a hillside on the left and another time behind a long shelf on the right. I think the robot should notice that an astronaut is not visible and broadcast an alert.

At this point, more than 3 hours after suiting up, Abby and Brent were getting tired, and the rest of us were exhausted just following them, trying to take steady video and photographs, and with hardly a place or moment to rest. John and Ed had refilled the Platypus water supply for Abby and Brent, and John refilled one of my three bottles. The temperature was 83F, 7% humidity, and no wind. We did try one final experiment, having the robot track Abby (using absolute GPS), and this worked well for about ten minutes. But then the robot veered, heading directly for a large rock, away from Abby. Kim reported from the hab that the robot "thought" Abby was back the other way. So much for absolute GPS.

With an impending need to swap batteries again, we elected to call it a day. The JSC team had previously requested to end around 5 PM so they could begin the long process of getting the robot back into the truck. Nathan and Rob teleoperated the robot out of the canyon. I filmed some final scenes of Abby and Brent walking (and sometimes dancing) around the bends, and up the cow path. They joked with British and southern accents about having cornbread tonight in Hanksville.

So although wiped out, we were elated, and all returned to the canopy-truck with broad smiles and cheerful comments about what we had accomplished. I thought back to the capability of the system a year ago-fixed on one spot at the head of the canyon, with huge gaps in network coverage, we were able to run 20 to 40 minutes, demonstrating only the most basic data logging and alerting functions. The robot was not involved.

Today we had moved 200 meters around several bends, with the robot towing a cart, carrying the sample bags, providing live video coverage with automatic tracking, taking photos and panoramas. The agent system did much more, allowing images to be grouped in collections, and automatically associating data with the named work site. For the astronauts, the roundtrip walk was about 1.5 km.

A few things didn't work -- the printer on the cart might have gotten too dusty and so no bag labels were printed, and a few commands such as moving the robot to a waypoint in the plan weren't tested. But easily we accomplished over 80% of the plan and most everything worked flawlessly. Relating back to my initial idea to return to MDRS this year and redo the Lith Canyon EVA, I was surprised to realize how much more had been accomplished.

We packed the SUVs, stowed the canopy, and after verifying that all was well getting the robot back to the truck, headed back to the hab. I began to prepare dinner about 7 PM (hamburgers, fried potatoes and onions, bean salad). The JSC team arrived about 7:30 PM and with surprisingly high energy we brainstormed about next year, how to get our proposals together, when we might meet to detail the major next steps. About 8:10 PM the Hanksville crew departed.

We talked a long time over dinner, half in the daze that comes from working in the desert for the day, half from the week of long days. We were all grateful to realize that tomorrow (Saturday) we would not have another EVA!

Plans for Tomorrow: Stow gear, pack Rick's truck going back to Ames; clean the hab.

Report Transmission Schedule: Final GreenHab and Engineering reports Saturday.

Maintenance: No problems.

EVA Narrative/Data/Interpretations: Pending reports

Inventory: Plenty.

Miscellaneous: As we turned to walk back, exhausted for the day, we saw the canyon continued far around the next bend, and unexplored terrain extended beyond, as far as the eye could see.